Approximately 25 percent of the 11 million Americans with diabetes suffer from peripheral neuropathy and diabetic related foot complications account for the majority of nontraumatic amputations of the lower extremity. In this population, osteomyelitis and deep infection of the foot are relatively common complications. Early diagnosis of these complications is crucial in the management of these patients, because prompt antibiotic treatment cure infection and therefore can decrease the rate of amputation. However, establishing the diagnosis of osteomyelitis is quite difficult in this setting because of concurrent conditions, such as peripheral vascular disease, cellulitis, neuropathy and osteoarthropathy, which can obscure the clinical manifestations of osteomyelitis. Much of the amputations are due to the lack of a single test that is highly sensitive, specific and cost effective in the early diagnosis of osteomyelitis. Current diagnostic tests including routine radiography, laboratory studies, nuclear medicine procedures, and magnetic resonance imaging (MRI) suffer from significant shortcomings such as inadequate accuracy and cost. It has been shown that [18F] fluorine deoxyglucose (FDG) and Positron Emission Tomography (PET) reveal sites of inflammation with high sensitivity and accuracy. In the pilot study conducted at our institution, we have been able to demonstrate that FDG-PET imaging has high accuracy in the diagnosis of orthopedic infection in a small patient population. The main objective of the proposed research study is to determine the efficacy of FDG-PET imaging in the diagnosis of osteomyelitis or deep infection in patients with diabetic foot in a large patient population. We also intend to compare FDG-PET imaging directly to MRI to determine whether FDG-PET imaging is superior to this commonly used technique. We will also examine the potential utility of a novel magnetic resonance (MR) technique (developed by one of our investigators) in detecting marrow edema and bone loss as a result of such complications. We plan to enroll 240 patients over four years for the purposes outlined in the application. By utilization of the proposed work, we will be able to demonstrate the sensitivity, specificity, and effectiveness of this technique in the management of patients with diabetic foot. We believe this promising technology has great potential for the accurate diagnosis of this serious and challenging clinical problem, and may substantially influence the outcome in these patients.